Tissue diagnosis of suspected lung cancer: selecting between bronchoscopy, transthoracic needle aspiration, and resectional biopsy.

In pursuing a tissue diagnosis of a suspected lung cancer, there is a range of procedures to choose from. The principal goals are ideally to diagnose and pathologically stage the patient's lung cancer at the same time, preferably by using the safest, least invasive, and least costly tests. If there is clinical or radiographic evidence of extrapulmonary spread of disease, including supraclavicular N3 nodal involvement or a malignant pleural effusion, then radiology-guided or open biopsy will confirm tumor cell type and stage the patient as unresectable. For patients with symptoms, such as increasing cough or hemoptysis, that are suggestive of airways involvement. with or without radiographic finding of central lesions, sputum cytology is the least invasive study with a high specificity. A positive finding of cancer is especially helpful if the patient is not a surgical candidate because of anatomic location of the lesion or severe physiologic limitations. The limited sensitivity of sputum cytology and poor NPV may improve with improved sputum induction and collection and processing techniques. Bronchoscopy with direct examination of the visible airways is most often the preferred invasive diagnostic procedure. Although the procedure should be geared toward sampling the highest staged lesion to provide an accurate tissue staging at the time of diagnosis, additional procedures can be performed in sequence to sample different nodal stations, is well as the primary lung mass. The incidental finding of an unexpected central airways lesions or a synchronous second endobronchial lung primary will also affect plans for treatment. Autofluorescence bronchoscopy can improve the sensitivity for detecting early intraepithelial neoplasia. Bronchoscopy for central and peripheral lung masses that are suspected to be lung cancer should be performed with ROSE whenever available. For visible endobronchial lesions, given the similar yield of EBBX and EBNA, EBNA may provide an immediate diagnosis, thus obviating additional, possibly morbid, procedures such as BB or EBBX. For submucosal lesions, EBNA is superior. For central cancers that are peribronchial, TBNA performed as for regional nodal sampling should have a yield that is comparable to TBNA for staging. TBBX and TBNA of peripheral nodules that are smaller than 3 cm have a lower diagnostic yield. Coming generations of thin bronchoscopes and improved radiographic guidance systems may improve our ability to biopsy these lesions with greater accuracy and safety. Under all circumstances, immediate cytology feedback with ROSE will confirm the adequacy of the retrieved specimen for a definitive tissue diagnosis, thus avoiding the need for extra biopsies, or worse yet, the need for a second invasive procedure because of insufficient diagnostic material. ROSE is educational to the clinician and fellow-in-training in getting immediate feedback on the procedural techniques and in learning pulmonary pathology, as well. The diagnostic sensitivity of TTNA is high, especially for the larger peripheral-based lung lesion, and TTNA is a relatively rapid procedure. TTNA's sensitivity falls for smaller or more central lesions, where the false negative rate can approach 25% to 30%; the risk of pneumothoraces and bleeding increases with central biopsies. Furthermore, TTNA usually does not provide information about nodal staging, unless the TTNA is initially directed toward central lymph nodes. The central airways are not examined in the same appointment to address issues of resection margins when there may be central spread of disease. TTNA should, therefore, be held in reserve for cases in which the sputum cytology and subsequent bronchoscopy are negative, and the patient is not a surgical candidate or refuses surgery, even if the cancer is potentially resectable. TTNA may then provide the tissue diagnosis to permit initiation of cytotoxic chemotherapy and radiotherapy. TTNA may also be helpful in cases where the likelihood of cancer is only intermediate, such that a specific benign diagnosis or an adequate sample without cancer will greatly reduce the likelihood ratio of missing a cancer, and justify to the patient and physician an approach of careful observation. To maximize the yield of these diagnostic procedures, there must be continued improvement in the hands-on teaching of clinical fellows and pulmonary practitioners in the use of the various techniques of TBNA and TBBX, as well as the applications of new endoscopic technology, such as EBUS. Definitive curative surgery remains the goal for patients with lung cancer, with accurate pathological staging performed intraoperatively. Complete lobectomy or pneumonectomy remains the standard resectional approach. Therefore, for patients with sufficient cardiopulmonary reserve who can be clinically staged as IA or IB, either by good quality CT with contrast or increasingly with 18-FDG PET, the initial tissue diagnosis may be at the time of surgery, when a frozen section preceding a complete lobectomy with lymph node sampling will combine diagnosis and therapy.